In past downturns, a plummeting rig count evoked simultaneous but quite different emotions—despair, as industry conditions spiraled rapidly downward in the near term; and hope, as the falling rig count sowed the seeds of a cyclical recovery for the long term. Better days lay ahead.

Today, the view is more complex, blurred by a crosscurrent of caveats, typically beginning with “Yeah, but ... .” For instance: Yeah, producers are dropping rigs, but they’re also high-grading their acreage. Yeah, the rig count has come down, but the industry is keeping its most efficient rigs.

Yeah, capex has been cut, but drilling and completion costs are also down, so E&Ps can do more with less. Just as important, today’s high-intensity wells are much more productive.

Equally difficult is forecasting the timing and impact on production of E&Ps’ various decisions to complete wells that were drilled but left uncompleted—so-called DUC locations—pending lower costs and better commodity prices.

Given these factors, how does a falling rig count balance out against higher productivity trends? And, are there grounds—other things being equal—for expectations of a slowdown in oil production growth such that crude demand and supply come closer to equilibrium, potentially resulting in stronger commodity prices?

Opinions vary. Recent data from the U.S. Energy Information Administration (EIA)’s monthly Drilling Productivity Report suggest the declining rig count may be contributing to lower levels of crude production by an amount that outweighs the impact of rising productivity. Its findings are based on an analysis of key unconventional basins: the Bakken, Eagle Ford, Niobrara and Permian, among the more oil-oriented basins, and the Haynesville, Marcellus and Utica, among the more gas-oriented basins.

A defining characteristic of the EIA methodology is an estimate of production related to new drilling activity. This is measured in terms of barrels per day (bbl/d) for each rig running in a given basin during a given month. For example, in the Bakken, this estimate rises from 576 to 592 bbl/d and then 610 bbl/d in March-April-May. The data in this progression are termed “monthly additions from one average rig” and are used by the EIA as a measure of rig productivity.

In essence, the EIA combines the latter with the number of rigs running in a basin each month to arrive at an estimated change in oil production related to new wells being drilled in a basin. An overall change in production for the basin is then estimated by netting out the contribution from new wells against the runoff of production related to legacy wells. The methodology thus strives each month to weigh changes in rig activity, adjusted for changes in rig productivity, against production declines in a basin.

Using the Bakken as an example again, the changing rig count being applied to the “monthly additions from one average rig” (576 bbl/d, 592 bbl/d and 610 bbl/d in March-April-May, respectively) results in estimated output from new wells of approximately 93,000, 72,000 and 61,000 bbl/d. With the decline in legacy production less variable at around 81,000 to 83,000 bbl/d per month, the net change in production from the basin moves from estimated gains of more than 11,000 bbl/d in March to declines of more than 9,000 bbl/d and upward of 22,000 bbl/d in April and May, respectively.

The bigger picture is that, if the EIA’s methodology is close to being an accurate estimate of future production levels, output from several of the oil-oriented unconventional basins are in the process of rolling over or have already done so. For example, the EIA data (adjusted by Global Hunter Securities to incorporate EIA revisions), shows production from the Bakken growing by more than 23,000 bbl/d in January of this year, but dropping by more than 22,000 bbl/d in May, as noted above.

Global Hunter Securities macro strategist Richard Hastings thinks that the worst of the crisis is past, suggesting a pathway to $60/bbl in second-half 2015.

Other key unconventional basins are showing a similar pattern. The Eagle Ford, which grew production by 33,000-plus bbl/d in January, according to the EIA data, is showing declines for the last two months, with estimated output down more than 32,000 bbl/d in May. Likewise, the Niobrara has shown production declining for the past two months, with estimated output dropping by over 14,000 bbl/d in May. This compares to an increase in Niobrara production of 9,000-plus bbl/d in January.

Gibson Scott, director of energy research with ITG Investment Research Inc., points to the “step changes” in both proppant intensity and fluid volumes being incorporated into Bakken wells.

The only oil-oriented basin to maintain a growth trajectory, according to the EIA data, is the Permian. Output in May grew by more than 10,000 bbl/d, but this was down sharply from the 44,000 bbl/d growth rate in January. Moreover, the rate of growth has fallen by approximately 10,000 bbl/d each month since February. Collectively, taking into account all seven basins under study (i.e. both oil- and gas-oriented), estimated output in May was down on a month-over-month basis by more than 55,000 bbl/d, compared to growth of almost 115,000 bbl/d in January of this year (see graphic).

Richard Hastings, macro strategist with Global Hunter, points to a variety of other sources that lend support for the trends indicated by the EIA study. In addition to various EIA and Department of Energy reports, he cited the North Dakota Industrial Commission’s February report, showing crude production down for the second month in a row, as well as data from the American Association of Railroads (AAR). The latter indicated a 14th consecutive week of contraction in crude and fuel carloads hauled in the U.S. and Canada.

“We’re flattening out and beginning to contract,” said Hastings, referring to production trends in the seven major unconventional basins. “And the big turnaround is being led by the Bakken. The worst of the crisis is likely behind us, strongly suggesting that the pathway to $60 per barrel in the second half of 2015 is now opening up, in our opinion.”

U.S. shale oil and gas production (across the seven major plays tracked monthly by the EIA) is going into reverse on a month-overmonth basis, based on the latest estimates for April and May 2015.

Risks to the model

The methodology used by the EIA has not gone without scrutiny by some industry observers, however.

For example, the question as to whether the EIA methodology adequately captures the impact of high-grading was raised in an April blog by RBN Energy LLC. In a review of the IEA’s March Drilling Productivity Report, RBN Energy identified as “one potential risk” the assumption of “average productivity rates for all rigs and all wells in a region.”

In a low commodity price environment, with E&Ps focusing on sweets spots using their most efficient rigs, “the productivity of their drilling activity inherently rises,” RBN noted. But because the report assumes “average productivity for all rigs in all new wells, it does not initially capture the full upside of the productivity gains of high-grading as it is occurring. Rather, the model assumes that retiring rigs have the same productivity as the remaining fleet, so it could overestimate production loss from retiring rigs, and at the same time underestimate the potential new production from the remaining active drilling rigs.”

Bottom line: If high-grading is happening by using the most efficient rigs, the risk is that the report “could predict production declines that might not happen,” said RBN. And if high-grading involves focusing on sweet spots, the report could miss the likelihood that “the initial production rates of new wells should exceed that of the average historical well.”

Completion step changes

In the array of factors contributing to greater well productivity, completion design is one of particular importance for wells in the Bakken, according to Gibson Scott, CFA, director of energy research with ITG Investment Research Inc., in Calgary, Alberta. Scott highlights “step changes” in both proppant intensity and fluid volumes being incorporated into Bakken wells.

“These are step changes in terms of the amount of materials we are putting into the well,” said Scott, noting that a typical Bakken well uses about 4 million pounds of proppant in its completion, but that a “proppant-intensive” well uses 15- to 25 million pounds. Similarly, a typical Bakken well using a cross-linked or hybrid fluid system would use 60,000 to 70,000 bbl of fluid, while a slickwater frack would require more fluid at a higher rate, climbing to 200,000 to 250,000 bbl.

“When we track the impact of the proppant intensity in the Bakken, we see in core rock a significant uplift in EURs [estimated ultimate recoveries] and NPV [net present value] of the wells,” he continued. “In some of the core regions, we see as much as a doubling of EURs and, factoring in the incremental well costs associated with that, we see as much as a three or four times uplift in the NPVs. However, when we track proppant-intensive wells in noncore rock, we see EUR uplift, but not to the extent that would justify the incremental cost to use those completion practices.”

“We’ve taken away the growth trajectory,” said Baird senior research analyst Dan Leben, “but we haven’t replaced it with a significant decline trajectory.”

Assuming the rig count in the basin remains roughly constant with early-April levels (down about 50% from peak levels), Scott expected Bakken oil production to roll over in July of this year and then gradually decline by roughly 5% through December of 2016. This assumes any DUC wells are tied into sales in line with historical rates (e.g. 10% tied into sales after two months, 50% after five months and 90% after eight months).

The energy research team at Robert W. Baird & Co. similarly focuses on well productivity, or “drilling efficiency,” against a backdrop of E&Ps’ moves to high-grade their acreage. Its projections of output from U.S. unconventional basins, which foreshadow “limited U.S. oil price upside,” incorporate a “near-term step-up in productivity, as efficiency-dilutive activity is shed (with rigs being dropped in noncore areas) and as drilling concentrates further into the best-returning, highest-producing acreage.”

In forecasting rig productivity, Baird models a 35% gain in unconventional basin productivity in 2015, followed by a baseline estimate thereafter of half the growth rate of the prior year. Using the Eagle Ford as an example, Baird cites EIA data showing a 2014 exit-rate productivity per rig of 631 bbl/d, representing an uplift of 128 bbl/d over the 2013 exit rate of 504 bbl/d. Translating this into a monthly rate of 11 bbl/d, Baird then factors in a 35% improvement in monthly rig productivity for 2015, followed by 50% of the prior-year gains for subsequent years. Thus, the estimated 15 bbl/d monthly improvement for 2015 is followed by a roughly 7 bbl/d monthly increase in 2016 and 3 bbl/d for 2017.

Baird projections shows unconventional crude output peaking from four key basins—the Bakken, Eagle Ford, Niobrara and Permian—in this year’s second quarter, at 5.4 million bbl/d (MMbbl/d) and then going into decline, dropping to 5.1 MMbbl/d by the end of 2015. Six months later, growth is forecast to resume, rising from 5.1 MMbbl/d at midyear to exit next year at 5.2 MMbbl/d.

With rig productivity expected to continue to “surprise to the upside,” the impact of the decline in the rig count on production levels is expected to be “materially limited,” according to Baird. Further, if just a 10% increase in rig productivity were to occur beyond its base case, Baird noted, any reduction in U.S. unconventional oil output due to a falling rig count would be erased.

In terms of prospects for crude prices firming in the wake of the recent rig count reductions, hopes are dampened—all else being equal—by offsetting rig productivity gains giving rise to ongoing oil oversupply, according to the Baird energy team.

“The challenge is that output was in excess of 5.1 million barrels per day when we got into oversupply,” said Baird senior research analyst Dan Leben. “Yes, we’re coming down, but we’re not coming down enough to solve the overall problem.

“We’ve taken away the growth trajectory,” he continued, “but we haven’t replaced it with a significant decline trajectory. That’s why we believe the oversupply problem in the U.S. is going to persist, barring significant geopolitical events, and that oil’s going to stay lower for longer than people expect.”

In terms of a continued flow of new production, Leben points to the Permian as having the most room to run. Its development timeline is about four years behind that of the Bakken or Eagle Ford, he noted, and were it not for a skew in productivity data related to vertical wells, Permian wells would compare very favorably because “a lot of the rock is just as good.”

The issue of deferred completions, creating a backlog or “overhang” of DUC wells, may be providing a misleading picture of production levels, according to Kathryn Miller, principal with Denver-based BTU Analytics. A slowing rate of production may not necessarily be reliably forecast, for example, if an inventory of locations with sunk drilling costs is waiting on the sidelines, capable of bringing on new production in short order once the decision to defer completion is reversed.

Kathryn Miller, principal with BTU Analytics, said that deferred completions could be contributing to a misleading picture of production levels.

While E&P managements often move with a herd mentality, Miller said, recent practices have been varied in deferring completions, which have tended to be undertaken by larger-cap companies. On its first-quarter conference call, Halliburton cited third-party sources for an estimate of up to 4,000 DUC locations; other sources suggest a range of 2,500 to 4,000. The eventual size of the DUC inventory, and when E&Ps begin bringing it down, will likely play the biggest role in month-to-month production estimates over the next three to four months, she said.

“These deferred completions can’t extend out too long,” said Miller. “The effect on pricing and returns will be too great if we let the size of the oilfield service workforce fall off. It’s not as if everyone can wait for a year and expect to complete wells, because if you’re an oilfield service provider you’re not going to keep them on for a year, and those people will find other jobs.”